Empty supermarket shelves are fresh in people's memories. Suppliers struggled to quickly meet unprecedented demand for personal protective equipment, hand sanitiser and toilet paper as consumers bought up everything from laptops and iPads to exercise equipment and home-improvement products. China hasn't helped: confidence has been shaken by its trade restrictions on barley, beef, coal, wine, wheat, cotton, rock lobster, timber, sugar and copper concentrate, with threats of more restrictions to follow.

But are these shaky supply chains the rule or the exception? Are they representative of the 5950 products Australians import each year from more than 220 different countries, or have we fallen into the trap of "selection bias" by focusing on examples that don't reflect the wider reality?

The Productivity Commission is endeavouring to find out. For its interim report on the vulnerability of supply chains it posed three questions about how many of our imports could be considered "vulnerable" and, of those, how many could be considered "essential" to the wellbeing of Australians. Although the results are preliminary, they offer insights into what the government should do next.

The first question was whether Australia is too dependent on any particular country for each of our imports. Specifically, the commission looked at whether the country-market for each product is "concentrated," meaning that Australia's main supplying country accounts for 80 per cent or more of our supply.

The second question was whether Australia could obtain those imports from alternative suppliers should that country, or our relationship with it, get into trouble. In the commission's view, alternative suppliers are limited if the main supplying country accounts for more than 50 per cent of the global supply.

The third question was whether Australia obtains its concentrated imports from the main global supplier in a concentrated market.

So, how concerned should Australia be? The commission found a mere one-in-20 of our imports are "vulnerable" according to these criteria, or just 292 of the 5950 goods that it analysed. More importantly, it found barely any of these 292 vulnerable imports are "essential" to the wellbeing of Australians, meaning they aren't essential products themselves or aren't used in producing essential goods and services. Examples include festive decorations, champagne, clothing and toys.

The commission's analysis shows most of Australia's supply chains are not a cause for concern. Indeed, the vast majority of them have performed well during the pandemic. The speed with which clothing manufacturers switched to making face masks and gin distilleries switched to making hand sanitiser are powerful examples of markets doing what they do best: allocating resources to where they are needed.

These findings have two important messages for the government as it focuses on making supply chains more resilient.

First, the old adage "if it ain't broke, don't fix it" should be front of mind. The government should only intervene in a supply chain if it can clearly identify a market failure that is producing an unacceptable risk. The Productivity Commission has shown this is rarely the case; and, even when there are risks, most of them are best left to the businesses that have skin in the game and much more experience in managing risk.

Relying significantly on a single country for Australia's imports, for example, is not a problem if alternative suppliers are available. Even when alternative suppliers don't exist, risks may still be low if new suppliers can quickly enter the market. And even if alternative suppliers aren't available and high barriers to entry act to slow or deter new entrants, risks may still be low if that particular good has close substitutes or if it is not essential in the first place.

Second, the commission's analysis shows there are significant costs if the government intervenes in supply chains unnecessarily. After all, global supply chains aren't an accident: they emerged as businesses sought to boost productivity, increase innovation and reduce the cost of living, benefiting the most vulnerable Australians the most.

Facilitated by improvements in technology, reduced transport costs and freer trade, these supply chains have allowed businesses across the world to be more specialised and achieve economies of scale. Unwinding them means unwinding these benefits.

While the benefits of government intervention in supply chains are uncertain at best, the costs of government intervention are not. By intervening unnecessarily, governments risk damaging one of the few sources of productivity growth at a time when it has flatlined in most countries, thus increasing the cost of doing business, increasing the cost of living, weakening innovation, and hurting jobs and growth. As with everything in economics, there are no free lunches.

Worse still, governments might actually make supply chains riskier. On-shoring supply chains might sound like a safer option, but if it means replacing a diverse set of international suppliers with a single domestic supplier then we make our supply chains more concentrated, more vulnerable and less reliable.

Sourcing the bulk of Australia's vaccine supply from the Australia-based AstraZeneca instead of a more diverse set of international options highlights the dangerous false sense of security that can come with "Made in Australia". Indeed, most of Australia's supply chain disruptions have historically been homegrown, ranging from bushfires, floods and cyclones through to business failures, unwise government policies and, more recently, the closures of state borders.

What should the government do if it identifies a credible supply chain risk? Bringing supply chains on shore and developing domestic capability tend to be the go-to options, but they are also the most expensive and potentially the least efficient choices. Other options include stockpiling sufficient reserves, developing contingent contracts, and seeking to reduce the barriers to entry and barriers to expansion in key markets - things the government should be doing anyway.

Bringing supply chains onshore may be politically popular, but the long-run costs are significant. Politicians advocating this policy need to carefully explain why it is necessary and, more importantly, be honest with Australians about the costs.